---
title: "Introducing Grok 4.7"
slug: introducing-grok-4-7
url: https://listedarticles.com/articles/introducing-grok-4-7
canonical_url: https://x.ai/news/grok-4-7
content_type: announcement
language: en
published_at: 2026-09-22T09:08:13.426Z
updated_at: 2026-09-22T09:08:13.426Z
author: "SpaceXAI"
authored_by: human
publisher: "SpaceXAI"
publisher_url: https://listedstartups.com/companies/xai
topics: ["AI", "LLMs"]
license: all-rights-reserved
word_count: 645
reading_minutes: 3
citation: "SpaceXAI, SpaceXAI. \"Introducing Grok 4.7.\" 22 Sept 2026. https://x.ai/news/grok-4-7 (all-rights-reserved)"
# The full text follows. The web page shows an extract and sends readers
# to the source above; quote the citation and link the canonical URL.
---

# Introducing Grok 4.7

> SpaceXAI announces Grok 4.7: what is new in the model release, where it improves, and how to access it—from the official x.ai news post.

Introducing Grok 4.7 | SpaceXAI[Back to news](/news)Sep 21, 2026

# Introducing Grok 4.7

SpaceXAI's most powerful model for coding and knowledge work. Twice as fast, at half the price of comparable models.[Try for free](/build?utm_source=website&utm_medium=referral&utm_campaign=grok-4-7-blog&utm_content=build-cta)[Start building](https://console.x.ai?utm_source=website&utm_medium=referral&utm_campaign=grok-4-7-blog&utm_content=build-cta)

Grok 4.7 is our most capable model for coding and knowledge work. It works longer on difficult tasks, checks its own work more carefully, and comes with our best-calibrated safeguards to date. Served at the same price and speed as [Grok 4.6](/news/grok-4-6), it is highly competitive in its class.
A scatter and line chart comparing Fable 5.1, Opus 5, Grok 4.7, GPT-5.6 Sol, and Sonnet 5 scores against average cost per task.55%CursorBench 4.0 score50%45%40%35%30%25%20%$18$15$12$9$6$3$0Average cost per taskFable 5.1Opus 5GPT-5.6 SolSonnet 5Grok 4.7CostTokensSteps

On CursorBench 4.0, which stresses longer-running coding tasks, Grok 4.7 is at the frontier in price-performance.

## [Model Improvements](#model-improvements)

Grok 4.7 uses a new, larger base model compared to [Grok 4.6](/news/grok-4-6). It was trained with a longer reinforcement learning run on a harder mix of tasks, weighted toward problems that take many hours to complete. The model is better at verifying its own work and managing longer context. We also trained Grok 4.7 to natively understand the [Grok Bot](/bot) harness, making it better at conversational tasks and general knowledge work.
Grok 4.7 xHighGrok 4.6 HighGPT-5.6 Sol MaxFable 5.1 MaxInput token price$ per million$2$2$4$10Output token price$ per million$6$6$20$50Software engineeringCursorBench 4.046.3%40.4%41.7%51.8%Software engineeringDeepSWE v1.171.0%*65.2%72.7%70.0%Electrical engineeringEEBench64.0%53.0%39.4%56.4%Multi-hour office workAA Briefcase v1.11,6571,5461,4871,678Multi-hour terminal workTerminal-Bench 4.038.0%20.3%37.3%57.9%Legal workHarvey Legal Agent Benchmark19.6%15.8%2.5%6.7%Clinical reasoningHealthBench Professional56.7%48.5%60.5%62.1%

* high effortToken prices and benchmark scores for Grok 4.7, Grok 4.6, GPT-5.6 Sol, and Fable 5.1. Benchmarks are CursorBench 4.0, DeepSWE v1.1, EEBench, AA Briefcase v1.1, Terminal-Bench 4.0, Harvey Legal Agent Benchmark, and HealthBench Professional. An asterisk on Grok 4.7 DeepSWE marks a high-effort score.

Grok 4.7 is better at creating documents and presentations. In GDPval and AA Briefcase, AI is asked to work on tasks done by professionals such as lawyers, nurses, and financial analysts. Grok 4.7 improves upon Grok 4.6 on both benchmarks and performs comparably to other frontier models.
Professional knowledge workMulti-hour office workElectrical engineering

Professional knowledge work

GDPval050010001500Elo score1735Fable 5.1 (max)1695Grok 4.7 (xhigh)1605Grok 4.6 (high)1542GPT-6 Astra (max)GDPval, AA Briefcase, and EEBench scores comparing Grok 4.7 with Grok 4.6, Fable 5.1, and GPT-6 Astra.

## [Safety & Cybersecurity](#safety--cybersecurity)

Grok 4.7 was built with an entirely new safeguard stack. It is the strongest model we’ve tested on refusals and jailbreak resistance. In dual-use domains like cybersecurity and biological work, it leads on both utility for benign tasks and safe refusal on dangerous ones, topping LatchBio’s biosafety benchmark at 62.4%.

Grok 4.7 balances strong cyber defense capabilities with low refusal rates for legitimate use. It shows the highest safety on HackerBench v0.3, our benchmark for risky and malicious cyber tasks, allowing only 3.3% of risky dual-use prompts through while rarely blocking legitimate security work. We’ve also started giving select cybersecurity partners invite-only access to Grok 4.7’s red-team capabilities for defense research.

## [Pricing and availability](#pricing-and-availability)

Grok 4.7 is available today in [Cursor](https://cursor.com) and [Grok Build](/build). It is also available through the [Grok API](https://console.x.ai?campaign=grok-4-7-blog&utm_source=website&utm_medium=referral&utm_campaign=grok-4-7-blog), third-party coding harnesses, and model routers and cloud platforms.

The model is priced starting at $2 per million input tokens and $6 per million output tokens. We also serve a fast variant with twice the output speed at twice the price.
[

Console

Create an API key](https://console.x.ai/team/default/api-keys?campaign=grok-4-7-blog&utm_source=website&utm_medium=referral&utm_campaign=grok-4-7-blog)[

docs.x.ai

Read the docs](https://docs.x.ai)

### Try it in Grok Build for free

Get started today at [x.ai/build](/build).`$ curl -fsSL https://x.ai/cli/install.sh | bash`